Continued misreadings of Piketty?

In VoxEu Odran Bonnet and others write "Thomas Piketty’s claim that the ratio of capital to national income is approaching 19th-century levels has fuelled the debate over inequality. This column argues that Piketty’s claim rests on the recent increase in the price of housing. Other forms of capital are, relative to income, at much lower levels than they were a century ago. Moreover, it is rents – not house prices – that should matter for the dynamics of wealth inequality, and rents have been stable as a proportion of national income in many countries."
They agree "...housing is not just wealth but also capital, and it contributes to wealth accumulation. However, to be in line with Piketty’s capital accumulation model, the measurement of housing capital must be based on actual returns to housing – that is, rent."
This is not clear to me. One can borrow on a house to invest and buy other houses. There are cases in Melbourne where people have about ten houses and flats by a young age mainly using housing to borrow more to buy other houses. The expenditure on the houses and mortgage can be deducted from the rent and the capital is used to essentially obtain more  capital without any extra effort or savings. This of course varies from place to place and the local rules.
The other misconception seems to that it does not immediately follow from r > g that inequality increases. The inclusion of housing also seems connected with Matt Rognlie critique but are not answered yet as Rajiv Sethi explained. But this conclusion also uses the fact that capital is unevenly distributed, as Matt Bruenig explains here. Another common error is discussed by Matt Bruenig recently.